EMRS: Liberals Crash But Labor Doesn't Pick Up The Pieces

EMRS Lib 36 (-6) ALP 31 (+1) GRN 15 (+2) IND/OTHER 18 (+3)
Election held now would deliver a very hung parliament
Rebecca White retakes Better Premier lead, most probably as a result of disapproval of Jeremy Rockliff's performance

An eagerly awaited EMRS poll is up and as you would have been nuts to not expect, the Rockliff Liberal Government has been harshly whacked over its recent collapse into minority and its pursuit of a controversial AFL stadium.  After over a year in which basically nothing happened in this quarterly series the Government has slumped six points to 36%.  I have this as the Liberal Party's third-lowest primary since February 2011 (when it was also at 36), beating only a 35 and a 34 polled during 2017, a time when the poll had serious issues that it later addressed with overestimating the Green vote.

For a long time I have been wondering what it would take to lift the Labor Opposition above the low 30s and put it on to numbers where it could at least push for being the largest party in the parliament, if not for majority.  Surely this would be it?  According to this poll, no!  Remarkably the poll finds Labor up just one point to a feeble 31% with the slack being picked up by Greens (up two) and independents/others (up three).  So is the latter a boost in support for the recent defectors Lara Alexander and John Tucker?  The seat by seat breakdown is extremely granular because EMRS in its dashboard presents the numbers as a share of the overall state vote (and the sample size is tiny anyway) but the combined Ind/Other share hasn't moved upwards in either the Bass or the Lyons sample (a lot of the gain was in Clark).  

If this was a federal poll, we would see Labor's 2PP up by, say, 4.5% in three months and Labor with a clear 2PP lead and on course to win off a low primary (similar to, say, the 2022 federal election).  But Tasmania's is not a 2PP system.  If your party primary vote is low then it hurts more, because it's so much harder to win seats from behind and because if the Green and Ind/Other votes go up enough then that becomes seats for those forces.   The Greens' 15% is their best reading since mid-2019 and the 18% for Ind/Others is I think the equal highest ever (matching the June 2022 reading which I suspect was contaminated by the federal election.)

The Better Premier poll sees Rebecca White take the lead from Jeremy Rockliff, 40-38.  Unusually by the standards of preferred leader polling, it's been fairly common for White to lead in the past (she often led against Will Hodgman and was also narrowly leading Peter Gutwein when he was a very new leader, until the pandemic arrived.) The early leads against Hodgman were because White was extremely popular in her own right at the time.  The leads against Gutwein were most likely because he was so new, but even so perhaps this suggests White remains well regarded in her own right (it's hard to say since no actual approval ratings for her have been seen since before the 2018 election).  The change in this one (with White up 4 points and Rockliff down 6) is, I suspect, entirely a negative reflection on voter concerns about Rockliff's performance.  If it were not so, why would Labor's primary have done so little?

What's happening with the government's polling is all too obvious and all too self-inflicted (there may well be relief it isn't worse) but what's going on here with Labor?   We don't have sufficient detail of any issues polling to go with this poll so it is speculation, but there is probably a more than niche perception out there that Tasmanian Labor is still to clean up all its internal problems and doesn't clearly stand for anything. Even though the party has thus far said quite clearly that it does not support the proposed stadium, there has been ducking and weaving about how the party will eventually vote on it in a Project of State Significance vote should such a thing occur.  The party seems to be more scared of being wedged than determined to advance a clear position.  The most obvious and clear voices against the stadium (the biggest issue in Tasmanian politics right now, and which it's safe now to say is unpopular off what is known of the commissioned part of this EMRS sample) are minor parties and some independents.

It should be noted that while the stadium and the loss of majority are the most obvious news cycle items at the moment, this is also the first EMRS poll since racing integrity and computer hacking scandals which saw Madeleine Ogilvie's performance as Minister criticised, and also conflict of interest claims against the Speaker Mark Shelton.  All this, together with the spate of resignations since the last election, plays into a narrative that this government's wheels are falling off and that it cannot be long for this world (if it even manages to complete this term without collapsing or feeling obliged to go to another early election).  

The problem with all this is that at the moment it is not clear in this poll that the government is losing.  If voters go to the polls with the Labor primary still trailing the Liberals, the government might be able to survive in minority, or alternatively it might be displaced by some sort of "rainbow coalition" of opponents that would probably be quite a mess.  If Labor is going to win, its time in government will be much easier if it can get its primary vote up to, say, 40, and find new pathways to govern and avoid having to rely on the Greens.  And if Labor doesn't get a wriggle on soon, it might get harder, because federal Labor's honeymoon cannot continue forever.  

This is only one poll - albeit one that performed well at the last election despite not even polling close to election day - so not too much should be read into it pending corroboration from any others.  I should also note again that EMRS is not a member of the Australian Polling Council, meaning that it does not publish extensive details of its methods.  (Cynics may also mention that it is now a member of the often conservative-linked C|T Group - unfortunately with uComms being union-linked, we don't get a lot of polling here that is free of some sort of connection.)  There are a raft of commissioned polls in the field concerning the stadium, and any that are even half-decent will ask voting intention questions (as also may some of those that are not that good).  Hopefully some more voting intention results will see the light of day during this process.  

Seat Estimate: Fools rush in ...

It is impossible to estimate what this poll would produce if an election was "held now" in the restored 35 seat parliament with any reliability.  The only things I think I can say with confidence are that the government would lose its majority, Labor would get nowhere near one, the Greens would gain seats (perhaps several, perhaps not) and it's very likely that somewhere there would be more independents or fourth party MPs elected than the one in 2021.  If the current poll's vote shift continues over a few polls then there should be enough information in the geographic breakdowns to see if there has been any big shift there.  

My very rough estimate of a seat conversion for this poll if it occurred at a 35-seat election is Liberals 13-16 seats, Labor 10-13, Greens 3-6 with 2-6 for Independents and others.  On current numbers the Liberals would not get a four-seat slate anywhere outside Bass and Braddon and would not exceed two in Clark.  The Greens would be unlikely to win in Braddon, but would be a shot at two in both Clark and Franklin (though I think the former would be a high chance of going instead to a second independent if one could be found to run).  Independent and minor party prospects are everywhere, but the problem is that this "independent" vote may include a degree of vote-parking from voters who will go back to the majors if no suitable independent candidate runs.  

April uComms

I hadn't previously covered off on an Australia Institute uComms robopoll in early April which at the time had primaries (with undecided redistributed) of Liberal 42.2 Labor 33.5 Greens 14.8 Ind/Other 9.6.  The poll also included a string of questions about fishing management issues that as usual primed and prodded respondents to express views about a low-profile issue they would not be often giving thought to; these results can be generally ignored.  The problem with drawing any conclusions whatsoever from this poll is that its numbers are all but identical to the same pollster and source's results in their April 21 2021 poll.  That poll was taken in the final fortnight of the state election campaign but was wildly inaccurate, underestimating the Liberals by 7.3% after redistributing undecided.  There had been evidence before that poll via benchmarking off EMRS that uComms was underestimating the government, but errors that large make a poll almost impossible to benchmark - the best that can be said of this year's uComms was that the government's result at the time was not bad (at least as good as surrounding EMRS samples, and perhaps in majority territory after adjusting for the behaviour of uComms.)

(uComms also releases 2PP data in its polls - just ignore it, it's a result of applying a standard polling format in a state it doesn't work in.)

That is, of course, all very ancient history now, but it will be interesting to see if any other polling comes out soon.